Our modeling went through four rough stages; the progression is illustrated in Figures 1 - 4. The complete Kiviat diagrams are described in the next sections.
Figure:
The Original Feature Model Space As Proposed
In the Simos Position Paper
Figure:
Another Stab at Spatial Representation, Combining
Zachman's Axes w/ Comparative Features
Figure:
Transitional Model. This model tried to tease out
how independent the elements are: A given representation
strategy should have a discrete location within the front
face of the cube, but a different value for intra- and inter-system
representation attributes.
The conclusion was that this was tending toward a hypercube
representation where we would need additional axes that would
be difficult to visualize graphically.
Figure:
Intitial Hand-Drawn Kiviat Diagram: This was
the first attempt on the flip-chart after the matrix was drawn
on the board. Note the spacing of the arrows. This was in
anticipation of addinig more attributes; instead the mood of the group
shifted to display the attributes already found in more
circular formats. Note the sequnce (draw counter-clockwise): Intra- and
inter, mapped to commonality, variability, and temporal. Later we
decided there were some ``dead junctions''. Also note that where most
of the arrows showed a spectrum ranging from implicit to explicit
support of the feature in question, the arrow labeled ``Representation''
had more qualitatively distinct intermediary points.